The other LLMs probably don't have the training data in the first place. | alt Hacker News

alt Hacker News

derrida • yesterday at 12:29 PM • 1 reply • view on HN

The other LLMs probably don't have the training data in the first place.

Replies

chrisjj • yesterday at 3:45 PM

Er...

"Study uncovers presence of CSAM in popular AI training dataset"

https://www.theregister.com/2023/12/20/csam_laion_dataset/.